Greetings,<br><br>We are interested in testing the main effects and interactions between 2 factors that have more than 2 levels (a 2x4 within subjects design) using the cluster-based permutation test for either time domain or time-frequency domain data. Has anyone done this using Fieldtrip scripts? There is only one such contrast for a 2x2 design thus the interaction effects can be tested using a t test on design cell differences. However, for a factor with more levels, is there a way to perhaps specify contrasts to test the full effect of the interaction? Would setting cfg.contrastcoefs to something work? Otherwise, is there a way to specify this in cfg.design?<br>
<br>Thanks so much for your time,<br><br>Bornali Kundu<br><br>